The present invention is for a method and system for pain classification and monitoring optionally in a subject that is an awake, semi-awake or sedated.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for monitoring a pain of a patient by analyzing at least two physiological signals, the method including comprising: a. obtaining said at least two physiological signals, comprising Photoplethysmograph (PPG) and Galvanic Skin Response (GSR); b. processing said at least two physiological signals to improve signal quality, thereby forming a plurality of processed physiological signals; and c. extracting from said at least two physiological signals at least three features thereby forming a first vector, wherein said at least three features are selected from a group consisting at least of: PPG mean Peak (P) amplitude, PPG peak to peak time intervals, PPG Peak-to-Peak High Frequency (PPG P-P HF) Power, GSR amplitude, and GSR peak to peak time intervals; d. transforming said first vector into a second vector, said transformation comprises normalization; and e. monitoring said pain of said patient by applying a classification algorithm adapted to classify said second vector into a graduated scale representing the level of pain, wherein said classification algorithm comprises an ensemble of classification and regression trees.
A method for monitoring a patient's pain level involves analyzing at least two physiological signals: Photoplethysmograph (PPG) and Galvanic Skin Response (GSR). The method includes obtaining PPG and GSR signals, processing these signals to improve signal quality, and extracting at least three features from the signals to create a first vector. These features are selected from: PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals. The first vector is then normalized into a second vector. Finally, a classification algorithm, specifically an ensemble of classification and regression trees, is applied to the second vector to classify the patient's pain level on a graduated scale.
2. The method of claim 1 , wherein said pain detection monitoring further comprises: communicating said monitored pain to a receiving unit selected from the group consisting of: a higher processing center, a person, a caregiver a call center, and any combination thereof.
The pain monitoring method, which analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, further includes communicating the monitored pain level to a receiving unit. This receiving unit can be a higher processing center, a person, a caregiver, a call center, or any combination of these.
3. The method of claim 1 , wherein said at least two physiological signals further comprise at least one signal selected from the group consisting of: electrocardiogram (ECG), blood pressure, respiration, internal body temperature, skin temperature, electrooculography (EOG}, pupil diameter, electroencephalogram (EEG), frontalis electromyogram (FEMG), electromyography (EMG), electro-gastro-gram (EGG), laser doppler velocimetry (LDV), partial pressure of carbon dioxide, and accelerometer readings.
The pain monitoring method, which analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, extends the analyzed physiological signals beyond PPG and GSR. It also includes at least one signal selected from electrocardiogram (ECG), blood pressure, respiration, internal body temperature, skin temperature, electrooculography (EOG), pupil diameter, electroencephalogram (EEG), frontalis electromyogram (FEMG), electromyography (EMG), electro-gastro-gram (EGG), laser Doppler velocimetry (LDV), partial pressure of carbon dioxide, and accelerometer readings.
4. The method of claim 1 , further comprising obtaining and processing a priori data, wherein said a priori data is selected from the group consisting of: environmental parameters, patient parameters, disease, stimulus, medicament and any combination thereof.
The pain monitoring method, which analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, also incorporates prior data. This a priori data can be environmental parameters, patient parameters, information about the patient's disease, the pain stimulus, medicaments being administered, or any combination of these factors. This data is obtained and processed.
5. The method according to claim 1 , wherein said ensemble of classification and regression trees algorithm comprises a random forest classification algorithm.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and classifies pain levels, the ensemble of classification and regression trees algorithm used is specifically a random forest classification algorithm.
6. The method according to claim 1 , wherein said ensemble of classification and regression trees algorithm comprises a Boosting framework.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and classifies pain levels, the ensemble of classification and regression trees algorithm used is a Boosting framework.
7. The method of claim 1 , wherein said ensemble of classification and regression trees algorithm is adapted for pain experienced with a particular disease, stimulus or medicament.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and uses an ensemble of classification and regression trees algorithm, the algorithm is adapted for pain experienced with a particular disease, stimulus, or medicament. This implies that the algorithm is trained or configured based on the specific context of the pain being monitored.
8. The method of claim 1 , wherein said patient is in a state of consciousness selected from the group consisting of: unconscious, under general anesthesia, sedated, partially sedated, awake, and semi-awake.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and classifies pain levels, the patient's state of consciousness is considered. The patient can be unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake.
9. A system for monitoring a pain of a patient by analyzing at least two physiological signals comprising: a. a signal acquisition module comprising a plurality of sensors and/or transducers for measuring and/or obtaining said at least two physiological signals from a subject, wherein said at least two physiological signals comprise Photoplethysmograph (PPG) and Galvanic Skin Response (GSR); and b. a processing module configured to process said at least two physiological signals, said processing comprising: i. processing said at least two physiological signals to improve signal quality thereby forming a plurality of processed physiological signals; ii. extracting from said plurality of processed physiological signals at least three features thereby forming a first vector, wherein said at least three features are selected from a group consisting at least of: PPG mean Peak (P) amplitude, PPG peak to peak time intervals, PPG P-P HF Power, GSR amplitude and GSR peak to peak time intervals; iii. transforming said first vector into a second vector, said transformation comprises normalization; and iv. monitoring said pain of said patient by applying a classification algorithm adapted to classify said second vector into a graduated scale representing the level of pain, wherein said classification algorithm comprises an ensemble of classification and regression trees.
A system for monitoring a patient's pain comprises: a signal acquisition module containing sensors and transducers to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals; and a processing module. The processing module: improves PPG and GSR signal quality, extracts at least three features to create a first vector (PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals), normalizes the first vector into a second vector, and monitors pain by applying an ensemble of classification and regression trees to classify the second vector into a graduated pain level scale.
10. The system of claim 9 , further comprising a display module for displaying said monitored pain.
The pain monitoring system, which comprises a signal acquisition module containing sensors and transducers to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module, which improves PPG and GSR signal quality, extracts features, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, further includes a display module. This display module is used for visually representing the monitored pain level to the user.
11. The system of claim 9 , wherein said signal acquisition module further comprises at least one sensor selected from the group consisting of: ECG, blood pressure, respiration, internal body temperature, skin temperature, EOG, pupil diameter monitoring, EEG, FEMG, EMG, EGG, LDV, capnograph, and accelerometer.
The pain monitoring system, which comprises a signal acquisition module containing sensors and transducers to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module, which improves PPG and GSR signal quality, extracts features, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, includes a signal acquisition module that has at least one sensor selected from: ECG, blood pressure sensor, respiration sensor, internal body temperature sensor, skin temperature sensor, EOG sensor, pupil diameter monitor, EEG sensor, FEMG sensor, EMG sensor, EGG sensor, LDV, capnograph, and accelerometer.
12. The system according to claim 9 , wherein said ensemble of classification and regression trees algorithm comprises a random forest classification algorithm.
In the pain monitoring system comprising a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module, which improves PPG and GSR signal quality, extracts features, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, the ensemble of classification and regression trees algorithm is specifically a random forest classification algorithm.
13. The system according to claim 9 , wherein said ensemble of classification and regression trees algorithm comprises a Boosting framework.
In the pain monitoring system comprising a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module, which improves PPG and GSR signal quality, extracts features, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, the ensemble of classification and regression trees algorithm is a Boosting framework.
14. The system of claim 9 , wherein said classification is adapted for pain experienced with a particular disease, stimulus or medicament.
In the pain monitoring system comprising a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module, which improves PPG and GSR signal quality, extracts features, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, the classification is adapted for pain experienced with a particular disease, stimulus, or medicament.
15. The system of claim 9 , wherein said patient is in a state of consciousness selected from the group consisting of: unconscious, under general anesthesia, sedated, partially sedated, awake, and semi-awake.
In the pain monitoring system comprising a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module, which improves PPG and GSR signal quality, extracts features, normalizes the data, and uses an ensemble of classification and regression trees to determine pain level, the patient can be in a state of consciousness selected from: unconscious, under general anesthesia, sedated, partially sedated, awake, and semi-awake.
16. The system of claim 15 , adapted to facilitate pain classification in an awake patient, a partially sedated patient or a sedated patient; wherein said first vector further comprises at least one feature selected from the group consisting of: respiration rate, skin temperature, accelerometer variability, EMG total power, pulse transition time and any combination thereof.
The pain monitoring system, which includes a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module which processes those signals, extracts features including PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels, and is adapted for unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake patient; is specifically adapted for awake, partially sedated, or sedated patients. In this case, the feature vector also includes at least one of: respiration rate, skin temperature, accelerometer variability, EMG total power, or pulse transition time.
17. The system of claim 15 , adapted to facilitate classification in a patient under sedation or under general anesthesia; wherein said first vector further comprises at least one feature selected from the group consisting of: EEG power of frequency bands, PPG dicrotic notch amplitude, pulse transition time, Coherence between 2 or more EEG/FEMG channels and any combination thereof.
The pain monitoring system, which includes a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module which processes those signals, extracts features including PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels, and is adapted for unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake patient; is specifically adapted for patients under sedation or general anesthesia. The feature vector also includes at least one of: EEG power of frequency bands, PPG dicrotic notch amplitude, pulse transition time, or coherence between two or more EEG/FEMG channels.
18. The system of claim 15 , adapted to facilitate pain classification in a patient under general anesthesia; wherein said first vector further comprises at least one feature selected from the group consisting of: PPG Peak-to-Peak (PPG P-P) variability Low Frequency (LF) power, PPG P-P variability Very Low Frequency (VLF) power, ECG R-R variability LF power, ECG R-R variability VLF power, pulse transition time, PPG dicrotic notch amplitude and any combination thereof.
The pain monitoring system, which includes a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module which processes those signals, extracts features including PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels, and is adapted for unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake patient; is specifically adapted for patients under general anesthesia. The feature vector also includes at least one of: PPG peak-to-peak variability low frequency power, PPG peak-to-peak variability very low frequency power, ECG R-R variability low frequency power, ECG R-R variability very low frequency power, pulse transition time, or PPG dicrotic notch amplitude.
19. The system of claim 15 , adapted to facilitate pain classification in an unconscious patient; wherein said first vector further comprises at least one feature selected from the group consisting of: EEG power of frequency bands, EEG peak frequency, Coherence between 2 or more EEG/FEMG channels, respiration rate, skin temperature and any combination thereof.
The pain monitoring system, which includes a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module which processes those signals, extracts features including PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels, and is adapted for unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake patient; is specifically adapted for unconscious patients. The feature vector also includes at least one of: EEG power of frequency bands, EEG peak frequency, coherence between two or more EEG/FEMG channels, respiration rate, or skin temperature.
20. The system of claim 14 , adapted to facilitate pain classification in chronic pain patients; wherein said first vector further comprises at least one feature selected from the group consisting of: respiration rate, skin temperature, accelerometer variability, pulse transition time, EMG total power and any combination thereof.
The pain monitoring system, which includes a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module which processes those signals, extracts features including PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels using an algorithm adapted for pain experienced with a particular disease, stimulus, or medicament; is specifically adapted for chronic pain patients. The feature vector also includes at least one of: respiration rate, skin temperature, accelerometer variability, pulse transition time, or EMG total power.
21. The system of claim 9 , wherein said first vector further comprises at least one feature extracted from patient monitors selected from the group consisting of: blood pressure, end tidal C02 level, end tidal level of anesthetic agents, body temperature, dosages of pain reducing drugs and any combination thereof.
In the pain monitoring system that acquires Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, processes them, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and classifies pain levels, the feature vector also incorporates at least one feature extracted from patient monitors. These features can be: blood pressure, end-tidal CO2 level, end-tidal level of anesthetic agents, body temperature, dosages of pain-reducing drugs, or any combination of these.
22. The method of claim 8 , adapted to facilitate pain classification in an awake patient, a partially sedated patient or a sedated patient; wherein said first vector further comprises at least one feature selected from the group consisting of: respiration rate, skin temperature, accelerometer variability, EMG total power, pulse transition time and any combination thereof.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels, and the patient is unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake; the method is adapted for awake, partially sedated, or sedated patients. The feature vector includes at least one of: respiration rate, skin temperature, accelerometer variability, EMG total power, or pulse transition time.
23. The method of claim 8 , adapted to facilitate classification in a patient under sedation or under general anesthesia; wherein said first vector further comprises at least one feature selected from the group consisting of: EEG power of frequency bands, PPG dicrotic notch amplitude and, pulse transition time, Coherence between 2 or more EEG/FEMG channels and any combination thereof.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels, and the patient is unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake; the method is adapted for patients under sedation or general anesthesia. The feature vector includes at least one of: EEG power of frequency bands, PPG dicrotic notch amplitude, pulse transition time, or coherence between two or more EEG/FEMG channels.
24. The method of claim 8 , adapted to facilitate pain classification in a patient under general anesthesia; wherein said first vector further comprises at least one feature selected from the group consisting of: PPG P-P variability LF power, PPG P-P variability VLF power, ECG R-R variability LF power, ECG R-R variability VLF power, pulse transition time, PPG dicrotic notch amplitude and any combination thereof.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels, and the patient is unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake; the method is adapted for patients under general anesthesia. The feature vector includes at least one of: PPG peak-to-peak variability low frequency power, PPG peak-to-peak variability very low frequency power, ECG R-R variability low frequency power, ECG R-R variability very low frequency power, pulse transition time, or PPG dicrotic notch amplitude.
25. The method of claim 8 , adapted to facilitate pain classification in an unconscious patient; wherein said first vector further comprises at least one feature selected from the group consisting of: EEG power of frequency bands, EEG peak frequency, Coherence between 2 or more EEG/FEMG channels, respiration rate, skin temperature and any combination thereof.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, classifies pain levels, and the patient is unconscious, under general anesthesia, sedated, partially sedated, awake, or semi-awake; the method is adapted for unconscious patients. The feature vector includes at least one of: EEG power of frequency bands, EEG peak frequency, coherence between two or more EEG/FEMG channels, respiration rate, or skin temperature.
26. The method of claim 7 , adapted to facilitate pain classification in chronic pain patients; wherein said first vector further comprises at least one feature selected from the group consisting of: respiration rate, skin temperature, accelerometer variability, pulse transition time, EMG total power and any combination thereof.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and classifies pain levels using an ensemble of classification and regression trees algorithm that is adapted for pain experienced with a particular disease, stimulus or medicament; the method is adapted for chronic pain patients. The feature vector also includes at least one of: respiration rate, skin temperature, accelerometer variability, pulse transition time, or EMG total power.
27. The method of claim 1 wherein said first vector further comprises at least one feature selected from the group consisting of: blood pressure, end tidal CO2 level, end tidal level of anesthetic agents, body temperature, dosages of pain reducing drugs and any combination thereof.
In the pain monitoring method that analyzes Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals, refines signal quality, extracts features like PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and classifies pain levels, the feature vector also includes at least one feature selected from: blood pressure, end-tidal CO2 level, end-tidal level of anesthetic agents, body temperature, dosages of pain-reducing drugs.
28. The system of claim 9 , wherein said acquisition module is further adapted to obtain a priori data selected from the group consisting of: environmental parameters, patient parameters, disease, stimulus, medicament and any combination thereof, and wherein said processing module is further adapted to process said priori data.
The pain monitoring system, which includes a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module which processes those signals, extracts features including PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and classifies pain levels, uses an acquisition module to also obtain a priori data. This data includes environmental parameters, patient parameters, disease information, stimulus details, and medication information. The processing module also processes this a priori data.
29. The system of claim 9 , further comprising a communication module for communicating said monitored pain of said patient to a receiving unit selected from the group consisting of a higher processing center, a person, a caregiver, a call center and any combination thereof.
The pain monitoring system, which includes a signal acquisition module to measure Photoplethysmograph (PPG) and Galvanic Skin Response (GSR) signals and a processing module which processes those signals, extracts features including PPG mean peak amplitude, PPG peak-to-peak time intervals, PPG peak-to-peak high-frequency power, GSR amplitude, and GSR peak-to-peak time intervals, normalizes the data, and classifies pain levels, further comprises a communication module. This module communicates the monitored pain level to a receiving unit, such as a higher processing center, a person, a caregiver, or a call center.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
May 14, 2010
August 20, 2013
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.